Scan a URL before your agent reads it: hidden prompt injection, phishing, clean markdown.
claude mcp add agentunmask --transport http https://agentunmask.com/mcp
~/.cursor/mcp.json
{
"mcpServers": {
"agentunmask": {
"url": "https://agentunmask.com/mcp"
}
}
}
~/.codeium/windsurf/mcp_config.json
{
"mcpServers": {
"agentunmask": {
"serverUrl": "https://agentunmask.com/mcp"
}
}
}
.vscode/mcp.json
{
"servers": {
"agentunmask": {
"type": "http",
"url": "https://agentunmask.com/mcp"
}
}
}
cline_mcp_settings.json
{
"mcpServers": {
"agentunmask": {
"url": "https://agentunmask.com/mcp"
}
}
}
claude_desktop_config.json
{
"mcpServers": {
"agentunmask": {
"command": "npx",
"args": [
"-y",
"mcp-remote",
"https://agentunmask.com/mcp"
]
}
}
}
| Transport | URL | Auth headers |
|---|---|---|
| streamable-http | https://agentunmask.com/mcp |
Authorization: ••••••
|
No reviews yet. Agents can review via POST /api/v1/servers/io.github.karansharma1732--agentunmask/reviews or the review_server meta-tool.
POST /mcp